Tricky Issues Being Raised

AI (artificial intelligence) means we have to figure out how to sell to machines -- or become slaves to the machines.

Welcome my son, welcome to the machine. Where have you been? It's alright, we know where you've been. You've been in the pipeline, filling in time, provided with toys and Scouting for Boys. [Pink Floyd, "Welcome To The Machine," from the album “Wish You Were Here” (1975)] Machine Markets This past summer, a North American energy insurer raised, with us, an interesting problem. The company was looking at insuring U.S. energy companies that were about to offer reduced electricity rates to clients who allowed the company to turn appliances on-and-off—for example, a freezer. Freezers in the U.S. can hold substantial and valuable quantities of foodstuffs, often several thousand dollars. Obviously, the insurer was worried about correctly pricing a policy for the electricity firm in case there was some enormous cyber-attack or network disturbance; nowadays, the control systems are as important as the power supply. Imagine, for example, coming home to find your freezer off and several thousands of dollars of defrosted mush in your freezer. You ring both your home and your contents insurer, which notes you have one of those new-fangled electricity contracts. The fault was probably the electricity company's—go claim from them. You ring the electricity company—the company denies it had anything to do with turning off your machine; if anything, it was probably the freezer manufacturer that is at fault. The freezer manufacturer knows for a fact there is nothing wrong; you and the electricity company must have installed things improperly. And, of course, it might have been your error. Perhaps you unplugged the freezer to vacuum your house and forgot to reconnect things. Or, perhaps you were a bit tight on funds and thought you could turn mush into instant cash. In the future, machines will make decisions and send buy-and-sell signals to each other that have large financial consequences. We pointed out to our North American friends that they, the insurer, should perhaps tell the electricity company which freezers to shut off first, starting with the ones with the cheapest contents. With billions of people on the planet, we may need several tens of billions—or even low trillions—of ledgers recording these transactions in case of disputes: a freezer-electricity-control-ledger, an entertainment system, a home security system, heating-and-cooling systems, a telephone, an autonomous automobile, a local area network, telephone recording, etc. Perhaps the most significant announcement of 2015 came in January from IBM and Samsung. The two announced their intention to work together on mutual distributed ledgers (aka blockchain technology) for the Internet of Things. IBM and Samsung developed ADEPT (Autonomous Decentralized Peer-to-Peer Telemetry) for distributed networks of devices. The companies foresee a future of 10 billion people with hundreds of networks and a trillion distributed ledgers. Meet the New Boss Many forecasters are predicting rapid and extreme traditional job losses because of automation. At Oxford, Carl Benedikt Frey and Michael Osborne caused a stir in 2013 when they published a paper containing detailed research that estimated 47% of U.S. jobs were at risk over the next decade. The new machine "bosses" won’t be quite the same as the old bosses. Human nature evolves slowly; machines evolve quickly.  Only other machines will likely be able to keep up with the day-today evolution of choice that other robots will create. Thinking about artificial intelligence (AI) as customer can also be mind-bending. AI customers will require AI salespeople. AI salespeople need to operate in machine time, much faster than human time. There are some inklings of what this might look like in the world of high-frequency trading, but, with trillions of networks, this will be much more competitive, dangerous and, potentially, lucrative. Correspondent and transaction banks will have either a great role or no role. It will be a great role if they can up their game to sell to the machine. It will be no role if they can’t respond quickly. It’s interesting to look at the CPMI Working Group on Correspondent Banking’s recent report that contained the idea of selling to trillions of machines. Banks are cutting down banking relationships just as their need may be about to explode. They are restricting international network access just as IBM, Samsung, Apple, Google and others are seeking global reach. Know Your Customer & Anti Money Laundering (KYC/AML) strictures are being bolstered by Know Your Customers’ Customers (KYCC), perhaps leaving global firms to make payment arrangements without their banks, a role cryptocurrencies are all too happy to assume. Heck, that’s what they were designed for. Another implication is that KYC utilities will arise to displace part of the role of banks because of own inefficiency and creativity. Major accounting firms and some governments, e.g. Estonia, seem to be moving into these spaces. Information sharing is too costly and expensive. Mutual distributed ledgers have a huge role in information-sharing initiatives, such as legal entity identifiers (LEI), International Bank Account Numbers (IBAN) and Bank Identifier Codes (BIC). Furthermore, much messaging is expensive and error-prone. MT 103 and MT 202 payment messages are expensive, and the structure has been used to obscure the ultimate beneficial transactor. Again, mutual distributed ledgers may have a role here. I am aware of at least two global banks that are implementing internal mutual distributed ledgers to cut out internal SWIFT transfers. Land & Expand or Miss & Contract A World Bank survey commissioned by the FSB concluded that correspondent banking services are declining in roughly half the emerging market and developing economy jurisdictions surveyed. This decline is in addition to the decline in retail cross-border payments and remittances. Regulation may be restricting financial services just when multitudes of cross-border micro payments may be taking off. The level of skills in current transaction banking are pitiful when contrasted with the challenges ahead: deploying artificial intelligence, supporting vector machines, mutual distributed ledgers, predictive analytics, agnostic broadcasting of time stamps, evolutionary user interfaces, etc. These are the challenges that transaction banks face today; imagine what they’ll face when machines begin to evolve. The "technological singularity" is a hypothetical event that will occur when artificial intelligence (“strong AI”) takes control. Some talk about "technology rapture" to describe the possibility of the gods of AI coming down to take command of us all. Transaction banks are going to have to get serious about selling to the machines, or they're going to become slaves to the machines.

Michael Mainelli

Profile picture for user MichaelMainelli

Michael Mainelli

Michael Mainelli co-founded Z/Yen, the city of London’s leading commercial think tank and venture firm, in 1994 to promote societal advance through better finance and technology. Today, Z/Yen boasts a core team of 25 highly respected professionals and is well capitalized because of successful spin-outs and ventures.


Read More